| According to the most recent report issued by the American Heart Association, cardiovascular disease remains the leading cause of death in the United States. Among cardiovascular diseases, coronary artery disease, which can cause, for instance, myocardial ischemia and myocardial infarction (or, a heart attack), accounts for 52% of the death toll. This underscores the urgent need and the critical role that a non-invasive, efficient and reliable imaging technique can play in the early diagnosis of coronary artery disease.;Echocardiography, namely cardiac ultrasound, has been extensively used in the clinic to evaluate both structural and functional changes of the heart due to its advantages of widespread availability, real-time capability, non-ionizing modality, low cost, portability and compatibility with pacemakers over other imaging modalities. The fact that abnormal myocardial motion and deformation (i.e., strain) is associated with coronary artery disease that leads to insufficient coronary blood and oxygen supply to the cardiac muscle (i.e., supply-type myocardial ischemia), has been well documented. Therefore, assessing myocardial motion and deformation may be the key to the detection of myocardial ischemia due to coronary artery disease.;Myocardial Elastography is such an ultrasound-based technique that utilizes cross-correlation on radio-frequency (RF) signals to estimate and image myocardial deformation in full echocardiographic views. Moreover, Myocardial Elastography aims at reliably identifying and localizing impaired myocardial segments attributable to coronary artery disease. Despite the fact that several ultrasound-based methods have been proposed to quantify 2D, or even 3D, myocardial motion and deformation, averaged temporal strain/strain-rate traces in a localized region are typically presented without simultaneously mapping motion and deformation images across the entire left-ventricle or the entire myocardial wall. In addition, a thorough fundamental performance assessment study of myocardial strain imaging has not been reported.;In this dissertation, Myocardial Elastography, a novel strain imaging technique, was evaluated in its full scope and firstly introduced, developed and evaluated using a theoretical framework based on a well-established, computational 3D model of the left ventricle and an ultrasonic image formation model. Full depiction of the nonuniformity of transmural (2D and angle-independent) myocardial deformation in short-axis views was shown. Not only were Myocardial Elastography strains obtained with good accuracy in comparison with the computational model solutions, but they were capable of differentiating the ischemic from the normal myocardial regions. In vivo validation of Myocardial Elastography deformation estimates in the canine left ventricle was thereafter performed against sonomicrometry under the conditions of progressive coronary blood flow reduction, which simulated the outcome of coronary stenosis and caused myocardial ischemia. Good correlation (r=0.84) and agreement (bias of 0.22% strain) of Myocardial Elastography with sonomicrometry were found. Most importantly, Myocardial Elastography was proven to detect the ischemic myocardial region at 40%, and possibly as early as 20%, flow reduction. Finally, a preliminary clinical validation study against MR tagging was conducted. Qualitative side-by-side strain image comparison and quantitative measures showing good correlation (r=0.75) and agreement (bias of 5.59% strain) in six normal and three pathological human left ventricles between the two imaging modalities were provided. The pathological myocardial regions were identified and characterized by the angle-independent strain estimates. In conclusion, the aforementioned findings collectively demonstrated the strong promise of Myocardial Elastography to constitute an early diagnostic tool for the reliable detection, localization and characterization of myocardial ischemia as a result of coronary artery disease. |
